【摘要】：太陽(yáng)能電池片生產(chǎn)中的質(zhì)量監(jiān)控系統(tǒng)對(duì)于光伏產(chǎn)業(yè)的發(fā)展至關(guān)重要。鑒于現(xiàn)階段多使用PC機(jī)作為機(jī)器視覺(jué)質(zhì)量監(jiān)控系統(tǒng)的處理控制中心,本文提出了基于嵌入式系統(tǒng)的太陽(yáng)能電池片缺陷檢測(cè)系統(tǒng),從成本、效率、功耗等方面對(duì)系統(tǒng)進(jìn)行優(yōu)化,本文完成的工作如下：(1)基于TI的達(dá)芬奇系列芯片TMS320DM6467構(gòu)建系統(tǒng)框架。使用ARM926和C6000系列DSP作為系統(tǒng)核心處理器。ARM負(fù)責(zé)對(duì)系統(tǒng)資源和外設(shè)進(jìn)行管理,DSP運(yùn)行算法對(duì)數(shù)據(jù)進(jìn)行處理。采用Codec Engine進(jìn)行雙核系統(tǒng)間的協(xié)作,DSP LINK進(jìn)行ARM和DSP通信。(2)針對(duì)太陽(yáng)能電池片生產(chǎn)過(guò)程中產(chǎn)生的缺角、斷極、斷柵、裂紋等缺陷,設(shè)計(jì)基于灰度差分圖像和行灰度分布圖的缺陷預(yù)判斷機(jī)制對(duì)缺陷進(jìn)行初步識(shí)別。排除那些可以確定不含缺陷的圖像,將資源集中在對(duì)那些可能存在缺陷的圖像檢測(cè)上面,以減小計(jì)算量,加快系統(tǒng)的檢測(cè)速率。(3)對(duì)經(jīng)預(yù)判斷機(jī)制檢測(cè)出的可能存在缺陷的圖像,根據(jù)電池片的幾何、輪廓等特征進(jìn)行更為精確、復(fù)雜的檢測(cè)操作,以確定圖像是否存在缺陷,并識(shí)別缺陷的基本類(lèi)型。(4)將算法按照符合C6000系列DSP的xDAIS標(biāo)準(zhǔn)和xDM標(biāo)準(zhǔn)進(jìn)行封裝。并移植到[MS320DM6467平臺(tái)上,實(shí)現(xiàn)了DSP端的檢測(cè)算法。編寫(xiě)ARM端應(yīng)用程序來(lái)管理系統(tǒng)內(nèi)存和IO接口等資源,最終實(shí)現(xiàn)基于嵌入式系統(tǒng)的太陽(yáng)能電池片的缺陷檢測(cè)。
[Abstract]:The quality monitoring system in solar cell production is very important for the development of photovoltaic industry. In view of the fact that PC is often used as the processing and control center of machine vision quality monitoring system, this paper presents a solar cell defect detection system based on embedded system, which optimizes the system from the aspects of cost, efficiency, power consumption, etc. The work accomplished in this paper is as follows: (1) the system framework is constructed based on the TMS320DM6467 of Da Vinci series chips based on TI. ARM926 and C6000 series DSP are used as the core processor of the system. Arm is responsible for managing the system resources and peripherals. Codec Engine is used to carry out ARM and DSP communication between DSPs and LINK. (2) aiming at the defects such as lack of angle, pole break, gate break, crack and so on, in the process of solar cell production, Codec Engine is used to realize the communication between ARM and DSP. The defect pre-judgment mechanism based on gray difference image and row gray distribution image is designed to identify the defect. Excluding images that can be identified without defects, and concentrating resources on the detection of images that may be defective, in order to reduce computation and accelerate the detection rate of the system. (3) for potentially defective images detected by the pre-judgement mechanism, According to the geometry and profile of the battery chip, more accurate and complex detection operations are carried out to determine whether the image is defective or not, and to identify the basic types of defects. (4) the algorithm is encapsulated according to the xDAIS standard and xDM standard of C6000 series DSP. And transplanted to [MS320DM6467 platform], realized the detection algorithm of DSP end. The application program of ARM is written to manage the memory and IO interface of the system. Finally, the defect detection of solar cell based on embedded system is realized.
【學(xué)位授予單位】：浙江工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類(lèi)號(hào)】：TM914.4;TP391.41
，

本文編號(hào)：2182158